search

dsriaditya999 / RGBXFusion

22 stars 3 forks source link

Why to set number of classes to 90? #5

Open mayone23 opened 6 months ago

mayone23 commented 6 months ago

I was going through the code, and it seems the model classify for 90 classes for flir dataset? Why so

connorlee77 commented 6 months ago

It's so we can use a coco-pretrained network without needing any single-modality fine tuning.

mayone23 commented 6 months ago

Okay. So for thermal and RGB, you are directly extracting features from the COCO-pre-trained network and then training with the fusion network, right?